i

Issues in Employment and Poverty

Discussion Paper

8

Labour Market Policies, Economic Growth

and Poverty Reduction: Lessons and Non-lessons from the Comparative Experience of East, South-East and South Asia

by

Rizwanul Islam

Recovery and Reconstruction Department International Labour Office, Geneva

April 2003

ii

Contents

Page Preface iii Author’s Acknowledgement v 1. Introduction 1 2. Recounting the Experience of Economic Growth and Labour Market Transformation in Countries of East and South East Asia 2

2.1 Growth and labour market transformation 2 2.2 Human capital and development 10 2.3 Labour market flexibility and social protection 11

3. The Asian Economic Crisis and a Reversal of Progress in Achieving

Labour Market Transformaton 12 3.1 The crisis, labour markets and poverty 12

3.2 Government responses to the crisis and issues for labour market policies 13 3.3 The economic recovery and setbacks due to the global economic slowdown 14

4. Economic Growth, Employment, Labour Markets and Poverty in South Asia 16 5. Lessons and Non-Lessons 26

References 31

iii

Preface

A high rate of economic growth is necessary for poverty reduction; but that is not sufficient. Employment is a major route out of poverty; for economic growth to be effective in reducing poverty, policies relating to employment and labour market, including human resource development, play an important role. The contrasting experiences of the East and South-East Asian countries on the one hand and South Asian countries on the other demonstrate this. Even though there were divergences in experience amongst the former group of countries, the general approach followed by them is quite influential in the thinking on development and poverty reduction. However, the adverse effects of the severe economic crisis faced by the countries of East and South East Asia during 1997-98 called into question a number of aspects of their model of development and poverty reduction on both economic and labour market fronts. The recovery achieved by them during 1999-2000 and the global economic slowdown of 2001 showed that market economies continue to remain vulnerable to periodic fluctuations. And while labour market gains erode quickly in a period of sharp economic downturn, recovery in labour markets follows economic recovery with a lag. The present paper presents a comparative analysis of growth and poverty reduction in countries of East and South East Asia and South Asia with a focus on employment and labour market policies. It also brings out possible lessons and non- lessons for the latter group of countries. Following are some of the major points that emerge from the paper. High rates of economic growth achieved by the countries of East and South-East Asia through outward oriented policies resulted in high rates of employment growth, and thus contributed to poverty reduction. Labour market policies in those countries emphasized human resource development and avoided generating rigidities. The result was a rapid transformation of the employment structure, vast improvement in the employment situations in general, and speedy decline in poverty. In contrast, the South Asian countries pursued inward- looking policies and achieved much lower rates of economic growth. Moreover, growth was much less employment- intensive compared to that in East and South East Asia. Labour market policies pursued in South Asia also neglected human resource development, and created significant rigidities. As a consequence, the transformation of the employment structure has been slow. Not surprisingly, the achievement in terms of poverty reduction has been less impressive. The economic crisis faced by East Asia and the experience with the period of recovery and further fluctuations have brought out a number of weaknesses of the East Asian model. First, although the countries of that region generally achieved impressive rates of poverty reduction, the issue of “vulnerability” of a large segment of the population needs to be addressed. Second, the post-crisis scenario raises the issue of sustainability of poverty reduction achieved during the period of rapid economic growth. The lag with which labour markets respond to economic recovery becomes an important issue. Indeed, the process of poverty reduction that is entirely dependent on labour market outcomes can remain vulnerable to external shocks. The importance of social protection and safety nets becomes important in that context.

iv

Coming to the issue of labour market interventions, it is important to see whether they really create rigidities that hinder economic expansion and employment creation. Indeed, it may be possible to adopt an approach where a combination of employment security and social protection with incentives for enterprises to invest create an environment of security and flexibility. April 2003 Rizwanul Islam Director Recovery and Reconstruction Department

v

Author’s Acknowledgement Assistance provided by Ms. Nina Torm in compiling the data presented in this paper is gratefully acknowledged. Comments offered by Dr. Ajit K. Ghose on an earlier draft helped in sharpening a number of the arguments made in the paper. Of course, the usual disclaimer applies. The views expressed in this paper do not necessarily reflect those of the organization with which I am associated. Rizwanul Islam

1

1. Introduction

Ever since the publication of the influential article on economic development by W.A. Lewis (1954), economic development in countries with surplus labour has been described as a process of economic transformation towards more modern sectors and the transfer of surplus labour from traditional sectors to modern sectors until such surplus is exhausted and real wages start rising. Indeed, development in countries of East Asia (viz., Republic of Korea; Taiwan, China; Hong Kong; and Singapore) during the 1970s, 1980s and the first half of 1990s followed that pattern. A number of countries of South East Asia, especially Malaysia and Thailand, and to a lesser extent Indonesia and Philippines were also on a similar journey towards the so-called “Lewisian turning point” – at least until they were hit by the economic crisis during 1997-98. The broad contours of development in the countries of East and South-East Asia (henceforth referred to as ESEA) mentioned above are quite well-known by now. On the economic front, they all followed relatively more outward-looking and export-oriented strategies. High levels of private domestic investment – financed in large part through high levels of domestic savings, coupled with investment in human capital acted as major engines of growth in all these countries (albeit in varying degrees). Labour markets were characterized by a high degree of flexibility in terms of the responsiveness of wage rates to changes in demand for labour and the ability of firms to hire and fire without too much intervention by either governments or trade unions. And it has often been argued that such flexibility in the labour market offered an environment in which enterprises could concentrate more on investment, economic growth and expansion of employment. The model of economic development described above became so influential that it was widely regarded as something to be emulated by other developing countries. However, the adverse effects of the severe economic crisis faced by the countries of ESEA during 1997-98 called into serious question a number of aspects of this model of development on both the economic and labour market fronts. Studies on the social dimensions of the Asian economic crisis demonstrate the serious adverse effects of the crisis on the unemployment, underemployment and poverty situations in the affected countries, and show that even commendable achievements in poverty reduction may prove unsustainable in the absence of social protection and safety nets for the workers (Betcherman and Islam, 2001; Lee, 1998). One could, of course, argue that the adverse effects of external shocks like the one inflicted by the Asian economic crisis are likely to be short- lived, and labour markets (and hence the poverty situation) should recover as economic growth resumes. Indeed, the economies of ESEA did start recovery fairly quickly (although the timing and strength of the recovery varied). But the global economic slowdown of 2001 showed that market economies continue to remain vulnerable to periodic fluctuations, and that in an increasingly integrated global economy, individual economies (especially the ones which are more export-oriented) can be quickly affected by developments elsewhere. And while labour market gains can erode quickly in a period of sharp economic downturn, recovery in labour markets follows economic recovery only with a lag. On the other hand, the countries of South Asia followed a different path of development, at least till the early or mid 1980s. They were generally more inward looking, and less export oriented (except, perhaps Sri Lanka). They appear to have paid less attention to human

2

capital and skill development. And their labour markets are said to have been characterized by less flexibility in terms of the ease with which firms could hire and fire their workers. While only small segments of their labour force are in the organized sector and are covered by union activities, it is usually believed that the ability of unionised enterprises to adjust to changes in economic and operating environment has been less compared to enterprises in countries of ESEA. The performance of the countries of South Asia in terms of economic growth and poverty reduction has been less impressive than that of ESEA countries. On the labour market front, they have remained much farther away from the coveted Lewisian turning point than ESEA. In terms of the effects of external shocks, the South Asian countries were able to avoid the contagion effects of the Asian economic crisis, but have not been spared by the global economic slowdown of 2001. In fact, segments of the modern sector labour market in Bangladesh have already felt the negative fallout of recession and slowdown in the developed countries. And as large parts of these economies (even within the ‘modern/organized’ sectors) remain outside the scope of social protection measures (e.g., unemployment insurance) a sudden shock leads to the reversal of whatever gains have been made on poverty reduction front. Employment acts as a major route out of poverty, especially in developing countries. And there are studies showing that during their period of high economic growth in the countries of ESEA, employment expansion did play a major role in poverty reduction (Islam 2001a; ILO, 2002b; Khan, 2001). However, the experience during the Asian economic crisis has shown that even employment- intensive growth cannot completely eliminate the vulnerability of the poor to external shocks unless there are in place adequate measures of social protection and safety nets (Lee, 1998;). As the East Asian experience of high growth and poverty reduction is often mentioned as a model for other developing countries to pursue, it is important to take a careful look at that experience as it unfolded during the period of high growth as well as of the recent economic crisis and subsequent recovery, so that any lessons that might be drawn can be based on a full account of the relevant perspective. The main purpose of the present paper is to present a comparative analysis of growth and poverty reduction in countries of ESEA and South Asia with a focus on employment and labour market policies, and to bring out possible lessons and non- lessons for the latter group of countries.

2. Recounting the Experience of Economic Growth and Labour Market Transformation in Countries of East and South East Asia.

2.1 Growth and labour market transformation

The first important point to note about ESEA is the high rate of economic growth for more than two-and-a-half decades from 1970 (Table 1). The exceptions were substantial declines in Indonesia and Malaysia during the 1980s and a sharp decline in Philippines during the 1990s. But both Indonesia and Malaysia were able to return to high growth paths during the 1990s.

3

Table 1: East and South East Asia: GDP Growth (% change per year)

Notes: a) Forecasts from ADB, Asian Development Outlook 2002. Sources: ADB, Asian Development Outlook 2002 for 1997-2003 World Bank, WDI 2002 For 1970-80, 1980-90 and 1990-96 The second remarkable feature of these countries is the important role of exports (as shown by figures of exports as a percentage of GDP presented in Table 2). In the cases of Indonesia, Malaysia and Thailand, the degree of export orientation increased very substantially during the 1970s, a trend which continued for Malaysia and Thailand during the subsequent decades. Philippines and Rep. of Korea (to be henceforth referred to as Korea) increased their export-orientation further form their high levels in 1970. All five countries achieved healthy rates of export growth during 1970-80. And after a slowdown for Indonesia and Philippines during 1980-90, all five achieved double-digit export growth during 1990-95. The export-oriented nature of these economies is thus clear from Tables 2 and 3. Table 2: East and South East Asia: Exports as a Percentage of GDP

Source: World Bank, WDI 2001 and 2002. Table 3: East and South East Asia: Growth rate of Exports (%)

Notes: a) ADB, Asian Development Outlook 2002 forecasts. Sources: ADB, Asian Development Outlook 2001 & 2002 for 1990-95 and 1996-03. World Bank, WDI 2002 for 1970-80 and 1980-90.

1970-80 1980-90 1990-96 1997 1998 1999 2000 2001 2002a) 2003a)

Indonesia 7.5 5.7 7.3 4.5 -13.2 0.9 4.8 3.3 3.0 3.6 Malaysia 7.5 4.8 8.3 7.3 -7.4 6.1 8.3 0.4 4.2 5.8 Philippines 5.9 5.0 2.8 5.2 -0.6 3.4 4 3.4 4.0 4.5 Rep. Korea 8.6 8.9 7.0 5.0 -6.7 10.9 9.3 3.0 4.8 6.0

Thailand 7.0 6.8 8.0 -1.4 -10.8 4.4 4.6 1.8 2.5 3.0

1970 1980 1990 1999 2000 Indonesia 13 34 25 35 39 Malaysia 14 33 29 122 131 Philippines 41 57 75 51 67 Rep. Korea 22 24 28 42 44 Thailand 15 24 34 58 66

1970-80 1980-90 1990-95 1996 1997 1998 1999 2000 2001 2002a) 2003a)

Indonesia 9.0 2.3 11.7 5.8 12.2 -10.5 1.7 27.6 -9.8 10.5 8.0

Malaysia 8.1 9.7 19.9 7.2 1.0 -7.3 16.8 17.0 -8.8 7.0 11.9 Philippines 6.5 2.9 14.4 17.7 22.8 16.9 19.1 9.0 -16.2 3.0 6.0 Rep. Korea 20.7 12.1 12.8 4.3 6.7 -4.7 9.9 21.2 -14 7.0 10 Thailand 8.3 12.5 18.7 -1.9 4.3 -6.8 7.4 19.6 -7.0 4.0 9.0

4

A picture of the structural transformation of the economies of ESEA can be obtained from data on sectoral composition of GDP and employment presented in Table 4. It is clear that in all five countries the share of industry in GDP increased significantly already during the 1970s, and that of agriculture declined correspondingly. However, the share of industry continued its increase after 1980 only in Korea and Thailand. But the share of agriculture continued to decline in all five, indicating a continuous process of structural transformation of these economies. On the other hand, data on the sectoral composition of employment appear to point to a somewhat less clear picture of transformation of people’s livelihoods. In Korea and Malaysia, there was a sharp decline in the share of agriculture in total employment during 1980-95. But in Philippines and Thailand, the decline was less sharp. And Indonesia lies at the other extreme in this regard, where the share of agriculture started declining only in the 1990s. Data on the share of industry further illustrate the difference between the various countries. In Indonesia, for example, the share remained unchanged during the 1980s and increased during 1990-95. In Philippines, on the other hand, the share of industry in total employment remained virtually unchanged during the entire period of 1980-95. Thus, it is only in Korea, Malaysia, and Thailand that there have been clear and rapid transformation in the structures of economies as well as the livelihoods of people. Although in economies with a substantial share of traditional sectors (and very little or no unemployment insurance), figures on open unemployment do not necessarily reflect the true degree of tightness of the labour market, they do provide some indication of the behaviour of the labour market.

5

Table 4: East and South East Asia: Sectoral Composition of GDP and Employment (% of total)

Notes: a) and b) These figures are from ILO, Papers and Proceedings of the Fifth Meeting of Asian Employment Planners, and those marked a) represent manufacturing only. c) These figures represent 1965. Sources: World Bank, WDI 2002. ILO, Papers and Proceedings of the Fifth Meeting of Asian Employment Planners, 29-30 November, 1993, Bangkok, Thailand.

1970 1980 1990 1995 Agriculture Industry Services Agriculture Industry Services Agriculture Industry Services Agriculture Industry Services Indonesia GDP 44.94 18.69 36.37 23.97 41.72 34.31 20.42 37.64 41.94 17.14 41.8 41.06 Empl. 65.9 7.9 a) 26.2b) 55.9 13.2/9.3 a) 30.2/34.5 b) 55.9 13.7 30.2 46.1 18.7 35 Malaysia GDP 29.44 27.39 43.17 22.61 41.04 36.35 15.22 42.2 42.59 12.95 41.4 45.65 Empl. 50.5 11.4 a) 38.1 b) 37.2 24.1/15.5 a) 38.7/47.3 b) 26 27.5 46.5 20 32.3 47.7 Philippines GDP 29.52 31.65 38.83 25.12 38.79 36.1 21.9 34.47 43.62 21.63 32.06 46.31 Empl. 53.7 11.9 a) 34.4 b) 51.8 15.4/11 a) 32.8/37.6 b) 45.2 15 39.7 44.1 15.6 40.3 Rep. Korea GDP 27.08 29.48 43.45 14.84 39.89 45.27 8.51 43.11 48.39 6.19 43.2 50.61 Empl. 55c) 15 c) 37 c) 34 29 37 17.9 35.4 46.7 12.4 33.3 54.3 Thailand GDP 25.92 25.31 48.78 23.24 28.68 48.08 12.5 37.22 50.28 11.18 39.16 49.65 Empl. 79.2 4 a) 16.8 b) 70.8 10.3/7.9 a) 18.9/21.3 b) 64 14 22 52.1 19.8 28.1

6

Data presented in Table 5a indicate that by 1997 (i.e., before the economic crisis hit the region), all these countries (except Philippines) were able to achieve very low rates of open unemployment (ranging from 0.9% in Thailand to 4.7% in Indonesia). This, however, is not to suggest that all of them had achieved a complete transfer of the surplus labour from traditional to modern sectors where they are fully employed. In fact, data on underemployment (defined in terms of a time measure of working less than 40 hours a week) indicates continued existence of surplus labour at least in Indonesia and Philippines. In other words, they have still not reached the so-called Lewisian turning point. To what extent Thailand has moved towards that stage also remains an open question, as more than 50 per cent of that country’s total employment is still accounted for by agriculture; and the high degree of seasonal variation of employment in that sector is quite well known. Table 5a: East and South East Asia: Unemployment Rate (as % of Total Labour Force)

Notes: ∗ In Malaysia, unemployed is defined to include those who are not actively looking for jobs. This category of unemployed account for over half of the total number of unemployed. According to an official in the Manpower Department an unemployment rate of 3.2% could be considered as full employment in Malaysia. a) EIU forecasts, b) These figures are from Godfrey 1993.

Sources: ADB, Key Indicators 2002 for 1984 and 1997-2000, BPS Statistics Indonesia 1991, EIU Country Forecasts 2000, 2001 and 2002, Godfrey M. 1993 Labour Market Monitoring and Employment Policy in a Developing Economy : A Study of Indonesia, ILO, New Delhi. World Bank, WDI 2002 for 1980 and 1990. National Statistics Office Websites: Republic of Korea: http://www.nso.go.kr Philippines: http://www.nscb.gov.ph Malaysia: http://www.statistics.gov.my Indonesia: http://www.bps.go.id Thailand: http://www.nso.go.th

1970 1980 1984 1990 1997 1998 1999 2000 2001 2002a)

Indonesia 2.3 b) 3 b) 2 2.5 4.7 5.5 6.4 6.1 8.1 8.8

Malaysia∗ 5.6 5.8 5.1 2.4 3.2 3.4 3.1 3.6 4.2

Philippines 4.8 6.2 8.1 7.9 9.6 9.4 10.1 9.8 11.3 Rep. Korea 5.2 3.8 2.4 2.6 6.8 6.3 4.1 3.7 3.4

Thailand 0.8 2.3 2.2 0.9 3.4 3.0 2.4 2.6 3.8

7

Table 5b: East and South East Asia: Underemployment (as % of Total Labour Force)

Notes: a) This figure has been calculated from data on working hours and total labour force from BPS labours/employees situation by province in Indonesia 1982. b) This figure is taken from ‘Papers and Proceedings of the Fifth Meeting of Asian Employment Planners 29-30 November 1993’. c) Figures are from Dhanani 2002. For Indonesia, the definition of underemployment is the percentage of people working less than 35 hours a week. For Malaysia the definition of underemployment is the percentage of people working less than 30 hours a week. For the Philippines, prior to 1987 underemployed refers to employed persons wanting additional work; from 1987 onwards, the underemployment is defined as those who work less than 40 hours per week, but still desire to have additional hours of work. The 1980 figure represents 1987. d) For the Republic of Korea, the definition of underemployment used in this table is the percentage of people working less than 20 hours a week. Yet for 1997 and 1998 alternative figures representing the percentage of people working less than 35 hours a week are available being 7.3 and 9.3 respectively. For Thailand the definition of underemployment is the percentage of people working less than 39 hours a week. e) This figure has been calculated from data on working hours and total labour force from the National Statistical Office ‘Report of the Labour Force Survey Whole Kingdom (round 3) August 1990 and 1991’. Sources: Betcherman and Islam 2001, BPS Indonesia: Statistical Yearbook 1991, BPS Labours/Employees Situation by Province in Indonesia 1982, Dhanani 2002 ‘Strengthening the Indonesian Labour Market Information System’, ILO Papers and Proceedings of the Fifth Meeting of Asian Employment Planners 29-30 November 1993. Bangkok, Thailand, KILM 2001-2002, Philippines 1998 Yearbook of Labour Statistics, Thailand Statistical Yearbook 1991, Thailand National Statistical Office ‘Report of the Labour Force Survey: Whole Kingdom (round 3) August 1990 and 1991’. National Statistics Office of Indonesia: http://www.bps.go.id The performance in the area of poverty reduction has been very impressive, except in the Philippines (Table 6). In Indonesia, the incidence of absolute poverty declined from 40 per cent of the population in 1976 to less than 12 per cent in 1996.1 In Thailand, during a short span of six years (1990-96), the rate of poverty declined from 27 per cent to 11.4 per cent. Malaysia also achieved similar results, albeit from an already lower level. And in all these cases, high rates of economic growth were associated with high degrees of employment intensity (as demonstrated by figures of output elasticity of employment). In Indonesia, for example, employment elasticity in manufacturing increased from 0.3 in the 1970s to 0.8 in the 1980s, before declining in the 1990s (Islam 2002). Likewise in Malaysia, and Thailand, employment elasticity remained high throughout the seventies, and started declining in the 1980s (Khan 2001).

1 These figures are based on the “old poverty line”. Figures in Table 6 are based on the so-called “new poverty standard” – when in 1998, the poverty line itself was redefined, and figures for 1980 onwards re-estimated.

1970 1980 1990 1997 1998 1999 2000 2001 Indonesia 26.9a) 38.6b) 35.8 39.1 37.8c) 35.5c) n.a. Malaysia 7.3 7.9 n.a. n.a. n.a. Philippines 14.0 23.1 20.5 11.3 11.9 21.9 19.9 16.6 Rep. Korea d) 1.4 2.0 2.8 3.4 3.3 n.a. Thailand 18.0e) 7.6 13.7 n.a. n.a. n.a.

8

Table 6: East and South East Asia: The Incidence of Poverty 1980 1990 1996 1997 1998 1999 2000 Indonesia

• BPS a)

28.6

15.1

17.7

N.a.

24.2

23.5

17.6 b) (2002) Malaysia

• ADB c) • Govt. of Malaysia d)

19.9 (1984)

15.2 (1989)

8.5 (1995)

6.8

7.6

8.1

N.a.

Philippines • NSCB e) • ADB f) • World Bank g) • Balisacan h)

40.2 (1988)

41 (1985)

32.7 (1985)

39.9 (1991)

34.3 (1991) 26.9 (1988)

35.5 (1994)

N.a. 32.1 (1994) 23.4 (1994)

31.8 36.8 25.1

N.a. N.a. 27.8

N.a. N.a. 26.3

33.7 39.4 N.a.

Thailand • NESDB i) • Mahmood and Aryah (2001) j)

27

11.4 11.4

12

N.a.

12.9 12.9

15.9 N.a.

14.2 N.a.

Notes: a) The figures are based on the so-called “new poverty standard” based on the 1998 standard. According to this standard, the December 1998 poverty line was 96,959 rupiahs for urban areas and 72,780 rupiahs for rural areas. b) This figure is from ‘Employment Policies for Poverty reduction during Indonesia’s Economic Recovery’, EMP/RECON, ILO. c) These figures refer to the percentage of poor households and the poverty line used is RM33 per capita per month in 1970 prices. d) The poverty line used is RM33 per capita per month in 1970 prices, which by using the CPI as deflator translates into the following lines for the other years: 1984=RM 73.2 and 1989=RM78.9. Data for 1984 is from the Fifth Malaysia Plan and data for 1989 and 1995 is from the Seventh Malaysia Plan. e) These figures represent the proportion of families whose annual per capita income falls below the annual per capita poverty thresholds (PhP 11.319 in 1997 and PhP 13.823 in 2000) out of the total number of families. f) These figures represent poverty incidence as a proportion of the population and are based on the official 1997 poverty line of PhP 6.077 per month. g) These figures are based on the basic needs poverty line, which in 1997 was PhP 4.495 per month. h) These figures are based on poverty lines with uniform level of living implied and on expenditure as indicator of living standard. This approach requires (i) obtaining a reference food bundle satisfying the minimum nutritional requirement of 2.000 calories per person per day, (ii) adjusting this bundle for regional cost-of-living differences, and (iii) estimating the non-food component from the consumption patterns of households whose total expenditures (incomes) are just adequate for meeting the food threshold (though not actually preferring to allocate all these incomes to food) (for more detail see Balisacan 1996). i) These figures are from World Bank: Thailand Economic Monitor May 2002 based on an analysis of the Socio-Economic Survey 2000 for the National Economic and Social Development Board (NESDB) using poverty lines of 728 baht per person per month for 1996 and 911 baht per person per month for 1998. j) These figures are from

Mahmood and Aryah in Betcherman and Islam (2001) and use the same poverty lines as under i)

Sources: ADB, Key Indicators http://www.adb.org. Balisacan A.M. (1996) What is the Real story on Poverty in the Philippines? A Re-examination of Evidence and Policy. BPS Statistik http://www.bps.go.id. Govt. of Malaysia (1984) Mid-term review of the Fourth Malaysia Plan , 1981-1985. Govt. of Malaysia (1986) Fifth Malaysia Plan, 1986-1990. Govt. of Malaysia (1991) Sixth Malaysia Plan, 1991-1995 . Govt. of Malaysia (1996) Seventh Malaysia Plan, 1996-2000. ILO, Jakarta and Geneva (2002). ‘Employment Policies for Poverty Reduction during Indonesia’s Economic Recovery’. Mahmood M. and Aryah G. (2001) ‘The Labour Market and Labour Policy in a Macroeconomic Context: Growth, Crisis and Competitiveness in Thailand’ in Betcherman G. and Islam R. (2001) East Asian Labour Markets and the Economic Crisis: Impacts, Responses and Lessons, World Bank, Washington. National Statistical Coordination Board (NSCB) http://www.nscb.gov.ph. World Bank WDI (1998). World Bank (2001) Philippines Poverty Assessment Volume I: Main Report quoted in Soriano T.M. and Imperial G.S. (2001) Philippine Labour Market Trends, paper presented at the “Seminar on Labour Market Policies: Its implications to East and Southeast Asia” Manila, 2001. World Bank: Thailand Economic Monitor May 2002.

9

Table 7: East and South East Asia: Real Manufacturing Wage Indices

Notes: Base-year = 1990 (except for Thailand where the base-year is 1991 and Malaysia where there are two separate series, one starting at base-year 1990 and the other at base-year 1993). a) The indices are taken from ILO, KILM 2001-2002, based on UNIDO data for wages and LABORSTA for CPI b) The indices are taken from ‘Labour Market Dynamics in Indonesia – Analysis of 18 key indicators of the labour market’ and are based on the wage survey of workers (Survei Upah Buruh) covering real wages of production workers in manufacturing. c) For Malaysia, 1980 represents 1982. Indices for years 1982 to 1992 (base-year being 1990) are calculated from nominal wage data from the ILO Yearbook of Labour Statistics 1992 and 2000 and the data reflects average wages per worker in manufacturing as a whole according to the International Standard Industrial Classification of all economic activities (ISIC) Revision 3, or its former version ISIC Revision 2. Indices from 1993 to 2000 (base-year being 1993) are calculated from nominal wage data from Monthly Statistical Bulletin January 1999 and December 2001, Department of Statistics, Malaysia. This data is from the Monthly Manufacturing Survey conducted by the Department of Statistics, which covers 74 industries, based on the Malaysia Industrial Classification 1972. d) Indices for years 1980 to 1995 are taken from KILM 2001-2001 and indices from 1996 to 1999 are calculated from nominal wage data given in the ILO Yearbook of Labour Statistics 2001. e) Indices for the Republic of Korea and Thailand are taken from KILM 2001-2002, based on LABORSTA data both for wages and CPI according to ISIC 2. For all countries the CPI for individual years is taken from LABORSTA and the method used to calculate the real wage index is as follows: Real wage index = (Nominal wage index i/Corresponding CPI i) * 100, where the nominal wage index = (the nominal wage for year i/ nominal wage for the base year 0) * 100 Sources: ILO, KILM 2001-2002, P. B. Irawan, I. Ahmed and I. Islam (2000) ‘Labour Market Dynamics in Indonesia – Analysis of 18 key indicators of the labour market’, ILO Yearbook of Labour Statistics 1992, 2000 and 2001, Monthly Statistical Bulletin January 1999 and December 2001, Department of Statistics, Malaysia, LABORSTA.

1980 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Indonesia a) 85.5 100 105.7 121.8 121.1 117.5 172.4 171.6 Indonesia b) 100 101.5 107.6 111.1 116.9 127.5 95.1 97.5 Malaysia c) 86.5 100 104.3 110.1 100 101.5 117.4 134.3 146.9 138.3 146.9 167.2 Philippines d) 75.3 100 99.1 101.4 97.7 102.5 100.6 94.5 95.4 76.4 77.3

R. Korea e) 45.6 100 106.9 116.4 123.2 133.9 140.8 150.7 151.6 136.7 155.6 165.2

Thailand e) 55.7 96.2 100 103.9 104.3 101.4 113.2 117.9 120.5 119.9 110.8

10

Real wages in manufacturing increased (of course at varying rates) in all five countries of ESEA during the 1980s (Table 7). And except for Philippines, growth in real wages continued during the 1990s (till the Asian economic crisis hit them in 1997). The growth in real wages was supported, by and large, by healthy growth in labour productivity (Khan 2001). Real earnings also registered impressive growth in these countries, except in Philippines. It is thus possible to conclude that in countries of ESEA, high rates of economic growth led to impressive achievements in poverty reduction through the employment route. High degrees of employment elasticity and healthy growth in labour productivity and real wages led to high rates of growth in real earnings of workers.

2.2 Human capital and development Investment in human capital formation can play a major role in boosting economic development that benefits the poor. Increases in the productivity of the poor engaged in wage employment as well as self-employment enhance the prospects of labour-dependent households to be able to participate in and benefit from economic growth. And the principal policy to enhance the productivity of the poor is to endow them with education and skills (Khan 2001). The level of educational attainment in turn contributes to growth (World Bank 1993). Public policy in the countries of ESEA played an important role in promoting human capital formation. This, however, is not reflected so much in the share of national income devoted to education as in the difference in the allocation of public expenditure between basic and higher education. The share of public expenditure on education allocated to basic education has been consistently higher in ESEA than elsewhere (World Bank 1993). However, as they moved from their initial phase of growth based on unskilled labour to growth requiring higher-level skills, the required level of educational attainment also shifted towards higher levels. And there the performance varied between countries. In Thailand and Indonesia, for example, levels of enrolment at the secondary level were initially lower than in Korea and Malaysia, and continued to remain low. And it is well-known that in recent years a shortage of educated workers has already been felt in both Indonesia and Thailand. The gap between various countries with respect to the educational level of the workforce is reflected in the data in Table 8. In both Indonesia and Thailand, workers with secondary education remained well below a fifth of the total workforce in 2000. In contrast, in Malaysia, workers with that educational level accounted far more than a third of the total workforce already in 1980 and increased to over half by 2000.

11

Table 8: Percentage Distribution of the Workforce/Employed Population by Level of Education Country and level of education

1980 1990 2000

Korea a Primary Secondary

n.a. n.a.

47.8 38.1

14.9 43.6

Indonesia a No schooling Primary Upper secondary

20.2 26.1 14.4

15.0 36.1 11.5

7.9 38.2 17.9

Malaysia b

No schooling Primary Secondary

18.8 43.6 34.0

9.8 34.6 49.9

6.6c

26.4c

53.8c Philippines b

No schooling Elementary High school

n.a. n.a. n.a.

3.9 44.1 31.7

3.0d

39.9d

34.0d Thailand b

No schooling Elementary Secondary

7.4 80.1 8.2

5.4 78.2 10.9

3.6d

65.0d

17.8d Notes: a) Percentage of workforce, b) Percentage of employed persons, c) Figures for 1999, d) Figures for 1998 Sources: Indonesia: Figures for 1980 are from Biru Pusat Statistik: Labours/Employees Situation by Province in Indonesia, 1982. For 1990 and 2000, the figures are estimated by Islam (2002) from SAKERNAS, the national labour force surveys. Korea: International Labour Office: Key Indicators of Labour Market 2001-02. Malaysia: Department of Statistics, Government of Malaysia: Malaysia Yearbook of Statistics 1991 and 2000 . Philippines: Department of Labour and Employment: Philippines Yearbook of Labour Statistics (various years). Thailand: Ministry of Labour and Social Welfare. Thailand Yearbook of Labour Statistics 1991 and 1996. National Statistical Office: Report of the Labour Force Surveys (various years).

2.3 Labour market flexibility and social protection Apart from high rates of growth of employment, productivity and real wages, one other characteristic of the labour markets of ESEA that is often mentioned is their flexibility, both in terms of their ability to effectuate quantitative adjustments as well as to adjust through variations in real wages. Labour market flexibility with respect to wages and termination of employment is often mentioned as a positive factor enabling these countries to adjust quickly and smoothly to changes in the operating environment. This, however, has to be looked at along with the level of development of labour market institutions (e.g., workers’ organizations and the practice of collective bargaining) in order to obtain a proper understanding of how adjustments can be achieved with a minimum of adverse social effects in terms of loss of livelihoods and increased poverty. And in that context, the existence of social protection for workers also becomes important. It needs to be noted, however, that wage flexibility manifested itself not so much in the absence of minimum wage legislations as in the actual market outcomes. Indeed, as will be seen later (in Table 16), minimum wages are established by law in all countries of ESEA.

12

What was important is that despite such legislations, real wage movements demonstrated considerable degree of flexibility. 2 A similar observation can be made concerning regulations for the termination of employment. As Table 16 will show later, there were legislations covering period of notice and severance pay required for termination of workers. But such regulations were not of much consequence either because of the rapid growth of employment or because of the lack of effectiveness in implementing such regulations. In reality, the countries of ESEA were characterized by a high degree of labour turnover. And the employers had little difficulty in relocating labour away from declining industries (Khan, 1994). Labour institutions were poorly developed in the countries of ESEA. That is reflected in the low trade union membership in these countries, except in Philippines (Lee, 1998). Workers’ organizations were either weak or under firm government control (Khan, 1994). By 1997, Philippines was the only country out of the five being referred to in this paper which had ratified the ILO Convention on Freedom of Association (No. 87). In fact, the overall environment in some of these countries was not conducive to the development of free and independent trade union movement. As a result, they lacked the industrial relations system that could facilitate an effective participatory process of adjustments to sudden and severe shocks such as the economic crisis of 1997-98. The other aspect of labour market which is important for reducing the vulnerability of poor workers and ensuring safety nets for them in periods of structural adjustments and adjustments to sudden shocks is social protection. And in that respect, the countries of ESEA did not have much until the Asian economic crisis of 1997-98. In fact, till the early eighties they had very little other than protection from illness, workplace accident, disability and certain minimal provisions in the event of death. It was only in 1988 that a pension system was introduced in Korea (Khan 1994). And in 1997, Korea was the only country to have unemployment insurance, and that too with limited coverage. For the other countries, the only notable form of social protection was the state-run provided funds designed mainly to provide retirement benefits. And retrenched workers could withdraw their balance of accumulated savings from the fund. But the coverage of this was also limited; in 1997, this option was available to 12 and 16 per cent of the employed in Indonesia and Thailand respectively. The highest coverage was in Malaysia (48 per cent) followed by Korea (38 per cent). (Lee, 1998).

3. The Asian Economic Crisis and a Reversal of Progress in Achieving Labour Market Transformation

3.1 The crisis, labour markets and poverty

The achievements of the countries of ESEA in the areas of economic growth, employment, labour market transformation, and poverty reduction were seriously threatened by the economic crisis that hit them during 1997-98. As is well-known the immediate effect of the crisis was economic contraction in all these countries, although the severity of decline varied from country to country (Table 1). Economic recession affected the lives and livelihoods of people through its effects on employment and labour markets. In terms of the social impact of the crisis, it caused increases in unemployment, underemployment, and poverty; declines in

2 See Table 7 for data on real wages in countries of ESEA.

13

living standards, and an upsurge in social tensions that endangered the fruits of the rapid economic growth achieved during the earlier decades. Labour markets can respond to economic contraction (or for that matter, to changes in aggregate demand, in general) in a variety of ways, with responses from the demand side as well as supply side.3 In reality labour markets in countries of ESEAs adjusted to the economic cris is in a variety of ways: increases in open unemployment and underemployment, declines in real wages and earnings, sectoral shifts in employment with a reversal towards informal segments of the economies, changes in labour force participation, and migration (both within and between countries). The actual mix of these responses varied from country to country, the combination reflecting the overall level of development of the country as well as the characteristics of its labour market.4 Quantitative adjustment (in terms of increases in open unemployment) was most prominent in Korea and Thailand. For Korea, this response is easily explained by the fact that the country had the highest proportion of employment in industry (Tables 4 and 5). At the other extreme, open unemployment registered a small increase in Indonesia at the peak of the crisis in 1998. Real wages declined in all five countries, but the sharpest (and a dramatic) decline took place in Indonesia. In fact, at the time of the crisis, Indonesia’s economy and the labour market still demonstrated a high degree of dependence on agriculture and informal sector. And the labour market adjusted to the crisis through a variety of responses that included reverse movement of the labour force to agriculture and other informal segments, reverse migration to rural areas as well as increased pressure on international migration, and increased participation (especially by women) in the labour force (Islam, et al., 2001). The adverse effects of the crisis on employment and labour markets led to an increase in the incidence of poverty, thus marking the reversal of declining trends in poverty (Table 6). Again, Indonesia suffered the most dramatic increase, which in turn was due mainly to the very steep decline in real wages. But the fact that poverty increased even in Korea and Malaysia (albeit from low levels) shows how vulnerable a sizeable section of the population can remain even when economies achieve strong economic growth. And that in turn indicates the importance of social protection and safety net measures from the point of view of containing the adverse social effects of severe economic contraction seen in countries of ESEA during the crisis of 1997-98. As these countries had very little by way of social protection measures and had little preparedness to put in place effective safety net measures, the vulnerable segments of the population could not prevent themselves from a slide back into poverty.

3.2 Government responses to the crisis and issues for labour market policies Of course, the governments of the crisis-affected countries did not remain inactive; they responded with a variety of interventions which can be grouped into two broad categories: active labour market policies (ALMPs) and programmes, and measures designed to support incomes. Before the onset of the crisis, the countries of ESEA had rather limited experience with ALMPs – at least of the kind found in industrial economies. But once the crisis hit, the countries responded with a variety of different measures. Indonesia, Philippines and Thailand either introduced new direct employment creation schemes, including labour- intensive infrastructure construction programmes and credit schemes to promote self-employment or 3 For a discussion of the conceptual issues involved, see Layard, Nickell, and Jackman (1994). 4 Betcherman and Islam (2001) provides an analysis of these aspects.

14

small enterprises, or reinforced existing programmes. Korea also started public works programmes for job creation. All five countries had functioning institutions providing vocational training and employment services; and their responses to the crisis included interventions in skill training and job search services specifically for displaced workers.5 As far as income support measures are concerned, we mentioned earlier that Korea was the only country to have any form of unemployment insurance at the time of the economic crisis. And during the crisis, some reforms were undertaken to extend the coverage to smaller firms and a wider category of workers. But despite these reforms, only 12 per cent of unemployed workers were receiving benefits in mid-1999 (Betcherman and Islam, 2001; Lee, 1998). The other forms of income support included the savings in the state-run provident funds and severance pay; but neither of them provided any substantial means of income support. While ALMPs have been implemented in varying degrees in the region, the extent and manner in which the various elements of ALMPs were used to respond to the economic crisis also vary – both within and between countries. And so does their effectiveness. Effective and efficient active labour market programmes require considerable capacity to design and implement them. The objectives set for the programmes, the target groups that are intended to benefit are also important. The role and nature of ALMPs could also vary at different stages of development. So, what is relevant and has worked effectively in one context may not be so in another. Republic of Korea and Malaysia appear to have utilized their public employment service (PES) as well as training system in assisting retrenched workers. These countries also put in place programmes of job creation through public works programmes and promotion of micro and small enterprises. On the other hand, Indonesia and Thailand did not appear to have made any significant use of PES in responding to the crisis. Thailand’s use of ALMPs has actually been rather limited. Indonesia’s response has been focussed mainly on public works programmes and limited use of skill training. Philippines appears to have been somewhere in between these extremes, using a variety of instruments, albeit in varying degrees. The experience thus seems to warrant the conclusion that all the countries did not put in operation the entire range of potential instruments. The obvious corollary of this conclusion would be that the countries could be more active in utilizing ALMPs as instruments for responding to the adverse labour market effects of the crisis (Islam, et al., 2001).

3.3 The economic recovery and setbacks due to the global economic slowdown The countries of ESEA were able to attain economic recovery fairly quickly. With the exception of Indonesia, the rates of economic growth were quite substantial already in 1999. And in 2000, the economic crisis appeared to have become history. But in the wake of the global economic slowdown of 2001,6 the process of recovery seems to have suffered a setback. Data on GDP growth presented in Table 1 shows the extent of recovery achieved by the five countries of ESEA during 1999-2000 as well as the setback suffered by them in 2001. In 1999, recovery was indeed spectacular in Korea and Malaysia. And in 2000, all five countries were firmly on the path of economic recovery. But in 2001, there were reversals in

5 Betcherman and Islam (2001) provides a detailed account of the active labour market policies and programmes adopted by the countries of ESEA during the crisis. 6 See Islam (2001b) for an analysis of the employment implications of the global economic slowdown of 2001 and how countries responded.

15

all these countries. As with economic recovery, the rates of decline (i.e., setback in recovery) also varied – the worst affected being Korea, Malaysia and Thailand. In order to have a proper perspective on the reversal of economic recovery in the countries of ESEA, it is essential to understand the root causes of the global economic slowdown in 2001 and how ESEA countries were affected by that slowdown. The decline in the growth of world output in 2001 reflected a weakening of the three major engines of growth of the global economy, viz., the US, the EU region and Japan. The adverse effects of this slowdown spread to other regions through the channels of trade and capital flows. In 2001, the growth of world trade declined to 2.2 per cent from 7.9 per cent in 2000. Given the export-oriented nature of the ESEA countries and their heavy reliance on USA, EU and Japan for export markets, the adverse effects of a slowdown in the latter quickly affected them. In 1999 and 2000, the countries of ESEA benefited from the growth of world trade as reflected in the high growth of their exports. In 2001, export growth, and as a result, GDP growth for all of them suffered declines. Global integration and heavy reliance on markets in a few developed countries thus created a two-way impact on the countries of ESEA (Islam and Krishnamurty, 2001). How did the labour markets perform during the period of economic recovery (1999-2000) and the subsequent slowdown (2001).7 During the recovery, open unemployment declined only in Korea and Thailand. And yet, the rates of open unemployment remained at levels much higher than the pre-crisis levels. In Indonesia and Malaysia, unemployment continued to increase, and in the Philippines it increased in 2000 after a small decline in 1999. Underemployment levels remained very high despite some decline. On the whole, it thus seems that the labour market situation did not improve much during 1999-2000 despite economic recovery. And in 2001, Indonesia, Malaysia and Thailand experienced increases in unemployment (see Table 5). Underemployment had increased (albeit in varying degrees) in all the crisis affected countries in 1998. Unfortunately, comparable data for recent years are not easily available for all countries. Data for Indonesia and the Philippines indicate declines in underemployment both during 1999-2000 and 1999-2001 respectively. However, despite small declines, the levels of underemployment remain very high in both these countries. In terms of the sectoral distribution of employment, there was a substantial increase in the share of agriculture and a decline in the share of industry in Indonesia during the crisis period. In 1999, the trend got reversed for both the sectors indicating positive developments in the labour market. In 2000, however, there was movement back to agriculture, although the share of industries continued to rise. The decline that year was in the service sector. During 2000-2001, however, the shares of all three sectors moved in the right direction (i.e., declined for agriculture and increased for industry and services). But for the other four countries, 2001 was a bad year for industrial employment – with the share of this sector in total declining in all. It thus appears that with the global economic slowdown of 2001, not only did economic recovery of the countries of ESEA suffer a setback, their labour markets were also adversely affected. Real wages, however, fared better – except, of course, in Thailand, where except for a small recovery in 1999, the decline continued through 2001. In Indonesia, which had a suffered a

7 The following discussion draws on Islam and Krishnamurty (2002).

16

steep fall in real wages in 1998, there was substantial recovery in both 1999 and 2000, so that they went back fairly close to the pre-crisis level. In Korea, Malaysia and the Philippines, recovery in real wage rates continued through 2001. The experience of 2001 thus seems to offer some interesting contrasts with that of the earlier crisis period in respect of the manner in which the labour markets adjusted to the economic downturn. Of course, the downturn of 1998 was very sharp, and the labour markets adjusted through a variety of mechanisms. On the other hand, the economic slowdown of 2001 was less sharp compared to the earlier crisis; and the labour markets adjusted mainly through increases in unemployment, and especially declines in manufacturing employment. The services sector also appeared to have played a role, as employment in that sector recorded positive growth in 2001 in all the countries of ESEA. Real wage rates do not appear to have played a role in facilitating labour market adjustments in 2001. An important aspect of the social effects of an economic crisis is the impact on poverty. As mentioned earlier, the Asian economic crisis of 1997-98 has had a severe adverse impact on the poverty situation of the affected countries. The important question in this respect would be whether economic recovery succeeded in reducing poverty and putting the countries back on the path of poverty eradication. Although lack of up to date data on poverty makes it difficult to arrive at a firm conclusion in this regard, data compiled and presented in Table 6 does provide some picture. In Indonesia, the head-count ratio of the incidence of poverty increased sharply between 1996 and 1998, but declined thereafter, falling back to the level of 1996 by 2002. Thailand provides a different experience where the declining trend in the incidence of poverty did get reversed during 1996-98, and that trend continued till 1999 when the incidence of poverty increased further. Data on Philippines provided both by the Government (National Statistical Coordination Board) and the Asian Development Bank indicate an increase in poverty between 1997 and 2000. World Bank figures on the other hand indicate an increase between 1997 and 1998 and a slight decline in 1999 – although the level of 1999 is still higher than that of 1997. The weight of evidence thus seems to indicate that while Indonesia has been able to reach its pre-crisis level of poverty, Philippines and Thailand are yet to be able to do so. Several conclusions seem to follow from the above discussion. First, although the crisis affected countries were able to achieve economic recovery fairly quickly, that was not followed by a full- fledged recovery of labour markets. While the labour markets had adjusted to the crisis through increases in unemployment and underemployment, changes in the sectoral shares of employment, and declines in real wages, the opposite processes did not take place immediately (with the exception, perhaps, of wages) when economic recovery started. Second, while earlier gains in poverty reduction were eroded significantly during the crisis, during the recovery period the process of erosion appears to have stopped only in Indonesia. Third, the setback in economic recovery suffered by the crisis-affected countries in 2001 has led to a further deterioration in their labour markets (Islam and Krishnamurty, 2002).

17

4. Economic Growth, Employment, Labour Markets and Poverty in South Asia The contrasting experience of South Asia starts with economic growth during the 1970s, 1980s and the early 1990s (Table 9). The contrast is particularly strong for 1970-80 and 1990-95. These economies remained much more inward-looking (except Sri Lanka) till 1990 as is demonstrated by figures in Table 10 compared to those in Table 2. It was only in the 1990s that the economies of South Asia started opening up. In fact, export growth exceeded double digit figures during 1990-95 in Bangladesh, India and Nepal (Table 11). The contrast extends to the structure of the economies in terms of the composition of GDP and employment (Table 12). In the three countries mentioned above, agriculture accounted for more than two-thirds of total employment in 1995. Industry accounts for only a quarter of GDP in South Asia. Changes in the structure of these economies towards high productivity modern sectors have been slower than in ESEA. And that is reflected in generally higher incidence of poverty and a much slower progress in poverty reduction in South Asia (Table 13). Not only has growth been lower in South Asia, but the employment intensity of growth (as indicated by employment elasticities with respect to output) has also been lower compared to East and South East Asia during the latter’s initial periods of high growth (Khan 2001 and Islam 2001a). In this context, one needs to recognize that employment elasticity reflects the inverse of labour productivity. A rise in productivity would thus imply a reduction in employment elasticity. Therefore, raising employment elasticity in individual activities cannot be the objective because that would mean a further lowering of productivity in economies that may already be characterized by widespread low-productivity employment. But it needs to be added that as long as employment elasticity remains lower than unity, that would be a situation permitting simultaneous expansion of employment with an increase in productivity. Two further questions need to be raised in the context of levels as well as changes in employment elasticity. Regarding the level, the desirability of an elasticity of lower than unity has been mentioned above. How much lower than unity it should be (i.e., the right order of magnitude for the elasticity of employment) depends on the level of development and the relative factor endowment of the country concerned. The magnitude would also have a good deal of sectoral variation. The overall elasticity being a weighted average of sectoral elasticities, greater allocation of investment in more labour- intensive sectors and higher growth rates in such sectors could yield a situation where the overall employment elasticity increases (even with declining elasticities in some sectors). And the result could be higher employment growth with given GDP growth or employment- intensive growth. Regarding changes in employment elasticity over time, it is expected to fall gradually as a country becomes more developed and relatively less labour abundant. For example, a GDP growth of five per cent per annum and an employment elasticity of 0.6 – 0.7 would enable an employment growth of 3 – 3.5 per cent, which in turn should be in excess of the labour force growth in most developing countries. Hence, sustained growth of such magnitude over a period of time could enable an economy to complete the so-called “Lewis transition” (i.e., the absorption of ‘surplus labour’ in modern sectors) within a reasonable period of time. With lower employment elasticity, the required GDP growth would be correspondingly higher.

18

Table 9: South Asia: GDP Growth (Annual Percentage Growth Rates)

Notes: The data for 1980-90 and 1990-95 is from the World Bank, World Development Indicators 1997. Figures from 1996-2001 are from ADB, Asian Development Outlook 2002 and figures from 2002-2003 are forecasts from Asian Development Outlook Update 2002.

Sources: World Bank, WDI indicators 1997. ADB, Asian Development Outlook 2002 (and update 2002). Table 10: South Asia: Exports of Goods and Services as a Percentage of GDP 1970 1980 1990 1995 2000 2001 Bangladesh 8.31 5.66 6.25 10.92 13.98 15.40 India 3.57 6.15 7.27 11.23 13.95 13.62 Nepal 4.9 11.54 10.53 24.22 23.75 23.42 Pakistan 7.77 12.49 15.54 15.95 15.53 17.23 Sri Lanka 25.45 32.33 29.21 35.52 39.71 37.93 Source: World Bank, WDI 2002. Table 11: South Asia: Growth Rate of Merchandise Exports (% per year)

Notes: Figures from 1970-80, 1980-90 and 1990-95 are calculated from the World Bank, WDI 2002 data on exports in constant 1995 US$. Figures from 1996-2001 are from ADB, Asian Development Outlook 2002 and figures from 2002-2003 are forecasts from Asian Development Outlook Update 2002. a) This figure represents 1973-1980; b) These figures are from ADB Asian Development Outlook 2002. Sources: ADB, Asian Development Outlook 2002 and Asian Development Outlook Update 2002. World Bank, WDI 2002.

1970-80 1980-90 1990-95 1996 1997 1998 1999 2000 2001 2002 2003 Bangladesh 2.3 4.3 4.1 4.6 5.4 5.2 4.9 5.9 5.2 4.2 5.4 India 3.3 5.8 4.6 8.0 4.8 6.5 6.1 4.0 5.4 4.0 6.0 Nepal 2.8 4.6 5.1 5.6 4.9 3.3 4.4 6.1 5.0 0.8 3.5 Pakistan 4.3 6.3 4.6 6.8 1.9 2.0 4.2 3.9 2.6 3.6 4.5 Sri Lanka 4.3 4.2 4.8 3.3 6.3 4.7 4.3 6 -1.3 2.8 5.5

1970-80

1980-90

1990-95

1996 1997 1998 1999 2000 2001 2002 2003

Bangladesh 6.8 6.0 12.6 11.8 14 16.8 2.9 8.2 12.4 -7.4 5.0 India 7.1 5.3 12.3 5.6 4.5 -3.9 9.5 19.6 -1.0 7.0 9.0 Nepal 15.4a) 4.1 23.3 1.9 10.2 11.9 18.2 42.4 3.7 -15 5.0 Pakistan -0.9 8.3 7.7 7.1 -2.6 4.2 -10.7 8.8 9.0 2.2 10 Sri Lanka 0.5 4.3 8.8 7.6 13.3 3.4 -3.9 19.8 -12.8 7.0b) 15.0 b)

19

Table 12: South Asia: Sectoral Composition of GDP and Employment

Notes: The figures from 1980 and 1991/92 are from ‘Papers and Proceedings of the Fifth meeting of Asian Employment Planners’, 29-30 November, 1993, Bangkok, Thailand. The figures for 1970 and 1995 are from the World Bank, World Development Indicators 2002. For these years, industry represents manufacturing only. The figures for 2001 are from ADB Asian Development Outlook 2002. Source: ADB Asian Development Outlook 2002. ILO, ‘Papers and Proceedings of the Fifth meeting of Asian Employment Planners’, 29-30 November, 1993, Bangkok, Thailand. World Bank, WDI 2002.

1970 1980 1991/92 1995 Agriculture Industry Services Agriculture Industry Services Agriculture Industry Services Agriculture Industry Services Bangladesh GDP 54.6 5.8 36.7 49.4 14.8 35.8 36.8 17.7 45.5 25.3 14.7 51.1 Empl - - - 68.7 7.9 23.4 65.1 14 20.9 63.2 - - India GDP 45.7 13.8 33.8 39.6 24.4 36 32.5 27.3 40.2 28.4 17.8 43.7 Empl - - - 71.0 13.0 16.0 65 14.5 20.5 66.7 12.9 - Nepal GDP 67.3 3.7 21.2 61.8 11.9 26.3 54.3 18.0 27.7 41.8 9.5 35.5 Empl - - - 91.0 4.0 5.0 80.0 8.0 12.0 78.5 5.5 21 Pakistan GDP 36.8 16.1 40.9 30.6 25.6 43.8 25.7 26.0 48.3 25.9 17.1 49.6 Empl - - - 52.7 14.6 32.7 47.4 16.5 36.1 46.8 18.5 34.6 Sri Lanka GDP 28.3 16.7 47.9 26.6 27.2 46.2 21.8 28.6 49.6 23.0 15.7 50.5 Empl - - - 44.5 11.8 43.7 47.7 15.7 36.6 37.3 23.4 33.6

2001

Agriculture Industry Services Bangladesh GDP 25.1 26.2 48.7 Empl India GDP 24.3 26.8 49 Empl Nepal GDP 38.5 20.1 41.3 Empl Pakistan GDP 24.7 25.1 50.3 Empl Sri Lanka GDP 20.4 27.4 52.1 Empl

20

Data on real wages (Table 14) in manufacturing in the South Asian countries are difficult to interpret as different sources result in different pictures. Take India, for example. Both UNIDO data and figures from Goldar (2002) indicate increase in real wages during 1980-90. But the figures start diverging from 1992 onwards, after which the UNIDO data show consistently higher increases in real wages. For Bangladesh also, UNIDO data show higher increases in real wages than the data from the country’s government statistics during the 1990s. Thus, data from the national sources do not indicate much increase in manufacturing real wages in India during the 1990s and only a moderate increase in Bangladesh. For Pakistan, there is no indication of a rise in real wages. On the whole, it would, perhaps be reasonably safe to conclude that the countries of South Asia are much further behind those of ESEA from the so-called Lewis ian turning point. The other area of contrast between South Asia and ESEA is in the field of human capital. As can be seen from figures in Table 15, nearly half the workforce/employed population in Bangladesh, India and Pakistan are illiterate. And the percentage of those with secondary education is much lower compared to countries of ESEA (compare Tables 8 and 15). In terms of other indicators of human capital like adult literacy, mean years of schooling and the UNDP’s human development index, countries of South Asia (with the exception of Sri Lanka) compare unfavourably with those of ESEA (Khan, 1994).

21

Table 13: South Asia: Poverty Incidence

Notes: a) These poverty lines are based on the ‘Cost of Basic Needs Method’, computed as follows: First a food bundle corresponding to actual consumption patterns in the country and providing 2,122 kcal per day per person was chosen. Regressions were used to find the prices paid by the poor and given these estimates food poverty lines were computed as Zfk=GjPjkFj where Fj is the per capita quantity of food item j in the bundle and Pjk is the price of j in area k. The next step is to compute the cost of basic non-food needs; the household per capita consumption for household I is denoted by yi and food per capita consumption by xi. First in each area k the non-food expenditures zl

nk=E [yi-xi | yi=zfk] among households whose total consumption expenditures are equal to their regional food poverty line zfk (yi= zfk) were estimated. The non-food allowance zl

nk can be considered as a lower bound for the cost of non-food basic needs. Next, upper bounds for the cost of nonfood basic needs zu

nk =E [yi-xi | yi=zfk] were estimated as the non-food expenditures (in each area) among households whose food expenditures are equal to the food poverty line (xi=zfk). The third step consists of summing up the food and lower and upper non-food allowances to obtain the lower and upper poverty lines by area. In area k, the lower poverty line is defined as zl

k = zfk + zlnk and the upper line is zu

k = zfk + zu

nk. For example, the four highest levels of the upper poverty line are for the Dhaka SMA (area 1), other urban areas of the Dhaka division (area 2), urban areas of the Khulna division (area 9) and the Chittagong SMA (area 5). b) These figures represent head count ratios based on calculations from the unit level records of the 50th round (1993-94) and 55th round (1999-2000) of the National Sample Surveys on Employment Unemployment. c) Refers to percentage of poor households. To define a poor household first the households satisfying the following two conditions are filtered a) households belonging to lowest four per capita expenditure deciles b) households which spend more than 50% of their household expenditure on food. Next the households for which per adult equivalent calorie consumption is between 2475 and 2750 kilocalories are isolated and per adult average expenditure on food is computed for those isolated households. This average value is referred to as minimum required per adult equivalent food expenditure or MRAEFE. The households (i) which spend over fifty% of their household expenditure on food and (ii) for which the per adult equivalent food expenditure is below the value specified under MRAEFE are defined as poor households. According to this definition it is revealed that the minimum required per adult equivalent food expenditure is Rs. 743 per adult per month during 1995/96. Those households spending more than 50% of the expenditure on food and adult equivalent food expenditure is less than Rs 743 per adult per month (excluding non-food) are considered as poor households. d) These figures are from the Central Bureau of Statistics and are based on a poverty line of Rs

1983-84 1985-86 1988-89 1991-92 1995-96 1997-98 1998-99 2000 Bangladesh

• BBS Upper Poverty line a) • BBS Lower Poverty line a) • BBS, PMS f)

58.5 40.91 62.6

51.73 33.77

57.13 41.32

58.84 42.69

53.08 35.55

47(1996)

46(1997)

44.7(1999)

49.8

India • Sundaram and Tendulkar b) • ADB Key Indicators

45.89

35.57(1993)

31.48 26.1

Nepal • NHDR 2002 d)

42(1996/97)

41.5

40.1

38.1

Pakistan • FBS e) • Arif et al. • Ali and Tahir

23(1991)

26.6(1993) 27.2(1993) 28.1(1993)

29.3(1994) 27.4(1994) 27.9(1994)

26.3 29.6

32.2 35.2

Sri Lanka • Govt. Figures c)

30.4

26.7

22

(Notes to Table 13 cont’d) 4,404 per person per annum. e) These figures are calculated from the Household Income and Expenditure Survey (HIES) data set conducted by the Federal Bureau of Statistics (FBS). All of the poverty figures are predicated on a minimum caloric intake based definition of the poverty line. To this end, the expenditure needed to meet the cost of the food bundle that would ensure the minimum caloric intake is calculated; but to which is also added the average estimated non-food expenditure of households whose caloric intake is exactly at the defined minimum level. Food and non-food expenditure of households are then added up to arrive at the poverty line. The FBS study uses a variant to this methodology and regresses per equivalent adult total consumption expenditure against the estimated daily per-capita caloric intake to come up with the poverty line with the assumption that households that consume the minimum caloric requirement also meet their necessary non-food consumption needs. In practice both methodologies measure poverty on the same definition. All the studies in the table define the minimum caloric intake level per adult equivalent at 2550 calories. f) These figures are from Bangladesh Bureau of Statistics, Poverty Monitoring Survey (PMS), Summary Results, 1999. The poverty lines in the PMS use the food energy intake (FEI) method and refer to caloric intake of 2122 kcal/person/day in rural areas and 2112 kcal/person/day in urban areas. Sources: Bangladesh Bureau of Statistics (BBS): Household Expenditure Survey (HES). World Bank (1998): ‘Bangladesh from Counting the Poor to Making the Poor Count’. Govt. of Bangladesh (2002): ‘Bangladesh: A National Strategy for economic Growth and Poverty Reduction’. Sundaram & Tendulkar: (2002) ‘The Working Poor in India: Employment and Poverty Linkages and Employment policy Options’. UNDP: Nepal National Human Development Report 2001. Sri Lanka Ministry of Finance and Planning, Dept. of Census and Statistics http://www.statistics.gov.lk/Documents/bul_pdf/poverty.pdf. ADB (2002): ‘Poverty in Pakistan: Issues, Causes, and Institutional Responses’ http://www.adb.org/Documents/Reports/Poverty_PAK/chapter_2.pdf

23

Table 14: South Asia: Real Manufacturing Wage Indices

Notes: The base-year is 1990 for all countries and all sources except Bangladesh. a) The base-year is 1969-70 and the indices are the averages of the indices of the 4 centres, viz. Dhaka, Chittagong, Rajshahi and Khulna. b) The base-year is 1970, c) UNIDO estimates. d) The data is from ASI (Annual Survey of Industries) and wage rates cover all industries. e) The wage data is from UNIDO covering wage earners and salaried employees and the CPI from Laborsta. f) The wage data is from LABORSTA based on adminis trative reports covering wage earners and salaried employees and the CPI is also from LABORSTA. g) The wage data is from LABORSTA based on labour-related establishment surveys covering wage earners. The data is limited to Colombo and refers to March and September. In the table the bullet-points ‘UNIDO’ represent nominal wage data taken from the UNIDO database (ISIC 300 = total manufacturing) and calculated into real wage index according to the following formula:

(a) A nominal wage index (NRi) was first calculated for year (i) by taking the wage data for the base-year 1990 as the base (100) and expressing the value for year (i) as a percentage of the base value, by means of the following formula: NRi = (Wi/Wo) * 100 where Wo is the nominal wage for the base-year (1990) and Wi the nominal wage for year i.

(b) The real wage index (Ri) was then computed by dividing, for each year (i), the nominal wage index (NRi) by the corresponding CPI (Pi), by means of the following formula: Ri = (NRi/Pi) * 100

The CPI values are taken from LABORSTA General Consumer Price Index. For India the CPI was specifically for industrial workers. For Sri Lanka the CPI was limited to Colombo. For Bangladesh the CPI is from Bangladesh Bureau of Statistics (1998) Statistical pocketbook covering general prices. Sources: Bangladesh Bureau of Statistics (1998): Statistical Pocketbook. Centre for Integrated Rural Development for Asia and the Pacific (CIRDAP): (1997) quoted in Bangladesh; ‘From Counting the Poor to Making the Poor Count’ (1998), The World Bank. Goldar, B. (2002): Trade Liberalization and manufacturing employment: The case of India, Employment Paper 2002/34. Employment Sector, ILO. ILO, KILM 2001-2002, based on figures from LABORSTA, WDI, UNIDO,. LABORSTA CPI. UNIDO (2002) Industrial Statistics Database 3-digit level for nominal manufacturing wages (data collected from national statistics offices by UNIDO).

1980 1986 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Bangladesh

• UNIDO • BBS a) • CIRDAPb)

82(1983)

100 110 115

97.8 107 114

101.5 107 113

113 119

114 121

127.9 111 121

114 123

130.4 120

India • UNIDO • Goldar d)

72

76.5

82.7 92.1

100 100

95.8 95.1

105.8 96.6

107.7 98.2

114.2 103.4

132.6 107.5

128.5 103.6

129.9 102.5

122

130.6c)

135.1c)

Nepal • UNIDO • KILM e)

76.9

100 100

157.4 115.4

138.7 99.7

144.8 98.4

118.7 99.1

Pakistan • UNIDO • KILM f)

72.6

100 c)

100

106.1

64.3

74.5

100.8

99.1 88.4

88.6

Sri Lanka • UNIDO • KILM g)

60.5

106.4

81.9

100 100

112.3 109.1

109.8 102.1

165.3 101.9

176.5 108.3

177.4 108.5

163.4 97.4

147.7 97.8

156.4 93.7

102.3

24

Table 15: South Asia: Percentage Distribution of the Workforce/Employed Population by Level of Education Bangladesh 1989 2000 No education Primary or less Secondary/higher secondary

64.3 17.9 5.0

46.6 24.3 8.5

India 1972-73 1983 1993-94 Rural Illiterate Up to primary Secondary Urban Illiterate Up to primary Secondary

M 61.6 29.1 2.8

25.5 37.8 16.0

F 91.8 7.1 0.3

70.8 16.6 6.0

M 52.1 29.3 6.3

22.3 30.5 19.5

F 85.1 11.1 1.1

59.3 19.3 9.0

M 43.7 29.1 10.8

18.4 25.8 24.2

F 78.5 14.2 2.5

48.9 19.7 12.3

Pakistan 1981 1990 1992 Illiterate Literate Secondary or lower

68.9 31.1 21.1

64.7 35.3 23.2

59.6 40.4 26.0

Notes: M = male, F = female. Sources: For Bangladesh, Bangladesh Bureau of Statistics, Report of the Labour Force Survey, 1989 and 1999-2000. For India: Institute of Applied Manpower Research, Manpower Profile India: Yearbook 2000. For Pakistan: Ministry of Manpower and Overseas Pakistanis , Workforce Situation Report, and Statistical Yearbook 1992. South Asia also offers a contrast to ESEA with regard to interventions of various forms in the labour market (at least up to the period before the Asian economic crisis). Of course, it does not become so obvious from a simple comparison of the two groups of countries with respect to labour market regulations (e.g., minimum wage and employment security laws) and non-wage labour costs (see Tables 16 and 17). Data presented in these tables (based on ILO, 1997) seem to indicate very little difference in the existence of minimum wage laws, notice period and severance pay required for laying off workers. In fact, severance pay in the case of Indonesia and Thailand was higher than in Bangladesh and Pakistan. In terms of the percentage of labour force covered by trade unions, it is only India that stands out in comparison to countries of ESEA. In fact, it is very low in Bangladesh. Likewise, not much difference appears to have existed with regard to non-wage benefits, e.g., old age, disability and death benefit, sickness and maternity benefits, and unemployment benefits. Table 16: Labour Market Regulations in Selected Countries of Asia

Minimum wage Employment security Degree of unionization Country

Established by law

Percentage of average wage

Period of notice (days)

Severance pay (monthly wage)

Percentage of labour force

Bangladesh Yes n.a. 30 3.0 3 India Yes n.a. 14-30 0.0 24 Pakistan Yes n.a. 30 2.0 10 Indonesia Yes n.a. n.a. 4.0 n.a. Korea, Rep. Yes 24.1 30 n.a. n.a. Malaysia No n.a. 28-56 n.a. 15 Philippines Yes 58.3 30 3.0 12 Thailand Yes 61.7-75.8 n.a 6.0 1.6

Notes: The data refer approximately to 1995. The situation may have changed since then. Source: ILO (1997).

25

Table 17: Social Security and Non-Wage Labour Costs in Selected Countries of Asia

Old age, disability and death

Sickness and maternity

Work injury

Unemployment

Bangladesh N Y Y Y India Y Y Y N Nepal Y N Y N Pakistan Y Y Y N Sri Lanka Y N Y N Indonesia Y N Y N Korea, Rep. Of Y N Y N Malaysia Y N Y N Philippines Y Y Y N Thailand N Y Y N

Notes: Y = Yes ; N = No. These data refer approximately to 1995. The situation may have changed since then. For example, unemployment insurance was introduced in Korea when the country joined OECD. Source: ILO (1997). The difference, however, has to be looked at in terms of market outcomes rather than in terms of the existence of regulations. As mentioned earlier (in section 2.3), despite the existence of minimum wage legislations, real wages in countries of ESEA demonstrated a considerable degree of flexibility. And despite employment security regulations, labour markets demonstrated a high degree of labour turnover. It is in terms of these actual outcomes that the countries of South Asia offer contrasting experiences. Conditions of termination of jobs are also more rigid in South Asia. Social security coverage for the organized sector was introduced in South Asia (albeit for a tiny part of the workforce) ahead of ESEA. In the matter of trade union freedom, the record of South Asia has in general been better in South Asia.8 The other notable contrast in the growth experience of South Asia and countries of ESEA is the period of economic cris is that affected the latter and the post-crisis recovery period. It is well-known that South Asia was able to come out of the Asian economic crisis without being affected by the contagion effects of the initial financial crisis that sparked off a broader economic crisis.9 A comparison of the GDP growth figures for south Asian countries (Table 9) during 1997-98 with those of the countries of ESEA for the same period illustrates this contrast.10 Moreover, the contrast between the growth performances of the two sub-regions in the post-crisis period appears to be much less pronounced than it was during the period of high growth of ESEA. During 1999-2000, only Korea and Malaysia achieved GDP growth rates that are substantially higher than the South Asian growth rates. Although the countries of South Asia could come out of the Asian economic crisis relatively unhurt, the same cannot be said of the global economic slowdown of 2001. It is true that the countries of ESEA were more severely affected by the global recession than those of South

8 See Khan (1994) for a more detailed comparative discussion of labour market interventions in countries of South Asia and ESEA. 9 What is less known (or at least less understood) are the causes of this divergent experience. Some obvious explanations could perhaps be found in South Asia’s lesser dependence on external short-term capital (especially to finance investment in the construction sector), continued control on capital account, and greater control on the financial sector in general, etc. 10 Exceptions have been the declines in GDP growth in Nepal and Pakistan, but they cannot be attributed to the effects of Asian economic crisis.

26

Asia (and that is also natural, given the former’s greater openness to the global economy and dependence on exports); but the latter did not remain totally unaffected. GDP growth rates declined in all South Asian countries (except India) in 2001, and the decline continued in 2002. Figures for export growth (Table 11) tell the story further. From healthy growth rates in recent years, Bangladesh registered a negative growth of exports in 2002. For India and Sri Lanka, the negative growth came already in 2001. It is thus clear that with an increasing degree of integration into the global economy, South Asia can no longer remain immune to the fluctuations that take place periodically. While the impact of such fluctuations in exports (and hence in production activities) can vary from country to country and also between sectors within the country, the possible impact on labour markets and poverty situations is illustrated by what happened in the nascent export oriented garments industry in Bangladesh. Of course, firm official figures are not easy to get, but newspaper reports indicate that some 300,000 workers (out of a total of approximately one-and-a-half million) may have lost their jobs during 2001-02 due to decline in exports. Similar must have been the impact of negative growth in exports in India, Nepal and Sri Lanka. With very little social protection (in the form of unemployment benefits, severance pay, etc.) for such workers, and not much by way of alternative employment prospects, the possibility of the affected workers and their families relapsing into poverty cannot be ruled out.

5. Lessons and Non-Lessons

From the point of view of labour market transformation and poverty reduction, the first important point that emerges from the earlier analysis, especially based on the experience of pre-crisis ESEA, is the importance of high rates of economic growth of an employment-intensive nature. Policies needed to pursue such a growth path have been discussed extensively in the literature;11 and there is no need to repeat them here.

As far as labour market policies are concerned, the importance of human capital – as a factor contributing to growth as well as enabling the poor to participate in and benefit from growth – comes out clearly from the comparative analysis of the experience of South Asia and ESEA. Going beyond this general conclusion, it is essential to point out the relative importance of investment in various levels of education and skill training. While the rate of return in general and growth dividend are higher for investment in primary education at lower levels of development, the importance of higher levels of education and skills also comes out from the growth experience of some of the South-East Asian countrie s (notably, Indonesia and Thailand).

The issue of the sustainability of high growth and impressive rates of poverty reduction became important during the economic crisis faced by countries of ESEA and the subsequent period of their recovery. The vulnerability of even those who are not officially below poverty line to severe external shocks and sharp contractions in output has become apparent. Indeed, the experience of the last few years (exemplified by the Asian economic crisis, the subsequent crises in other parts of the world, and the global economic slowdown of 2001) shows that market economies continue to remain vulnerable to periodic economic fluctuations. And with a growing degree of integration into the global economy, the hitherto inward looking economies of South Asia have also become vulnerable to global economic fluctuations. The

11 See, for example, World Bank (1993) for a discussion on policies that created the necessary conditions for high growth in countries of ESEA. For more recent analysis of the East Asian experience, see Stiglitz and Yusuf (eds.) (2001), and Islam and Chowdhury (2000). Khan (2001) provides a good analysis of employment policies for translating the benefits of growth into poverty reduction.

27

need for social protection and effective safety nets for workers affected by such fluctuations can no longer be overemphasized. A variety of alternative means (ranging from unemployment insurance and severance pay for affected workers to various forms of active labour market policies/programmes) are in principle available for this purpose. And countries should look at the feasibility and applicability of those means in their respective situations.

On unemployment insurance, the conventional view is that such a scheme would not be feasible for developing countries. Lee (1998), however, makes a strong case for this form of income support. He addresses the technical issues of affordability and feasibility form the point of view of a number of countries and concludes that “an average required contribution rate of between 0.3 and 0.4 per cent of payroll from 1991 to 2000 would have made a significant contribution to cushioning the harsh impact of the crisis on modern sector workers” (Lee, 1998, p. 83). On the other hand, there is the view that many developing countries, especially those of South and South East Asia, have a small formal sector, and unemployment insurance may not be the most effective means of providing support to the unemployed workers. There are studies (Edwards and Manning, 2001, for example) suggesting alternative approaches that also merit careful consideration. What is important is to be alert about the possible fallout of an increasing degree of integration of countries into the global economy and periodic fluctuations therein. The recent example of large scale job losses in the small modern sector of Bangladesh points to the importance of preparedness to provide workers affected by such job losses with at least the minimum income support needed to prevent them falling back into poverty. The precise means of doing so has to be worked out in an undogmatic way after due consideration to available alternatives.12

A similar remark can be made about active labour market programmes. Given the concept of ALMPs and their application, especially in developed countries, it is clear that they provide instruments for labour market interventions that are meant not only for crisis or difficult economic situations. In fact, these are measures intended to facilitate the process of reintegrating the unemployed into the employed labour force. Even in the crisis-affected countries of ESEA, elements of ALMPs have been in operation already before the crisis – albeit in varying degrees in different countries. They were typically not viewed as ALMPs, but as policies and programmes to improve skills, create jobs, and place job-seekers, quite irrespective of fluctuations in outputs and employment. In view of the adverse labour market impact of the crisis, the countries attempted to ameliorate the situation by using them for responding to the crisis. It is thus important to develop a strong base and preparedness with ALMPs during normal economic situations so that they can also be applied to respond to

12 The above remarks, however, need to be qualified by several observations. First, the safety net and social protection mentioned here refers to only a very small proportion of workers because the organized sector in these countries represents a very small part of the economy as a whole. It is, however, important to give attention to this issue for at least two reasons. First, in an increasingly market-oriented and integrated global economy, the possibility of sharp economic downturns leading to widespread loss of employment is a real one. And unless safety nets for workers who may be affected are institutionalised, it would be difficult to prevent their relapse into poverty. Second, in the absence of institutionalised safety net measures the inevitable outcome is a crowding of retrenched workers into informal sector activities, thus creating a pressure on the already precarious situation of those originally dependent on them. Thus, even though the safety net mentioned here may cover only a small minority of workers, it could have a broader implication for poverty as a whole. It would, of course, be necessary to find measures of social protection for vast numbers engaged in the informal economy; and that is taken up later in this sector. Moreover, there may be questions concerning costs of and administrative capacity for implementing social protection measures – even for a small segment of workers. While these issues are beyond the scope of the present paper, it needs to be mentioned, however, that there are studies (e.g., Lee, 1998; Edwards and Manning, 2001) looking at various alternatives.

28

economic fluctuations and crises.13 In addition, continuous monitoring and evaluation of the efforts of ALMPs is crucial to keep an eye on developments and needs of labour markets. A strong labour market information system is essential in that context (Islam, et al., 2001).

Regarding the impact of labour market interventions on economic growth and employment, it is important to take into account the divergent views that exist. The conventional view is that institutional interventions in the labour market in the form of minimum wage laws, job security regulations, and measures of social protection cause high labour costs and thus reduce demand for labour. According to this view, they also impede adjustments to economic shocks by reducing flexibility in the labour market. The other view (Freeman, 1993, for example) raises doubts about the existence of excessively high and distortionary regulations in the labour markets of developing countries. This alternative view points out the broader benefits of labour market regulations which help improve productivity and foster economic growth, thus creating the necessary conditions for poverty reduction.

In judging the impact of labour market interventions, it is important to look at several things. For example, minimum wage laws need not necessarily create rigidities if the minimum is not a very high proportion of average wages, and remains flexible in reality. ILO (1977), reviewing the experience of a number of developing countries in this respect, concludes that minimum wages have indeed shown flexibility. That report also looks at various non-wage labour costs (due to various social security programmes like unemployment benefits, old age benefits, sickness and maternity benefits, etc.), and concludes: “none of the labour market policies traditionally emphasized by the distortionist view showed any of the hypothesized negative effects on employment and growth” (ILO, 1997, p. 188).14

Coming to the issue of labour market interventions in South Asia, one notices that non-wage and social security benefits basically consist of sickness and injury benefits, old age benefits, and the like. The employment security benefits include periods of notice and severance pay which don’t look much out of line with those in countries of ESEA. While it is true that such benefits are not available to vast numbers working outside the formal sectors in these countries, they can hardly be described as causing excessive rigidity and hindrance to hiring of labour. Indeed, they may even be characterized as modest levels of social protection that even a developing country should aim at. What is, of course, important is to see whether they reduce incentives for enterprises to invest, grow and hire more workers. The recent example of economic and employment growth in India (especially during the 1990s) can be illustrative in this context. India’s GDP growth increased from 3.5 per cent during the 1950s – 1970s to 6.4 per cent during the 1990s; and employment growth also increased from 0.68 per cent during the 1980s to 1.02 per cent during the 1990s. The incidence of poverty declined from 46 per cent in 1983 to 29 per cent in 1999.15 Yet, there is very little evidence to indicate that the labour market in that country has become more flexible (in terms of interventions either with respect to employment security, wages, or other non-wage costs) during the 1990s compared to the earlier decades. The real change that appears to have taken place is an

13 It needs to be mentioned in this respect the actual composition of the ALMPs has to vary depending on the overall level of development of a country, and how tight its labour market is. In countries of South Asia (and some South East Asian countries) where underemployment is still high and availability of employment is the main problem, the focus of ALMPs has to be more on job creation measures rather than job placement. 14 In fact, ILO (1977) refers to an econometric study (done under the auspices of the World Bank) which found that higher levels of social security contributions were associated with higher, not lower, total employment growth and had no significant effect on output growth in Latin America and the Caribbean. 15 Sundaram and Tendulkar (2002).

29

improvement in the environment for making investment and doing business in terms of a variety of regulations and restrictions.

Indeed, in a global economic environment characterized by periodic downturns, it is even necessary to put in place measures that can provide workers with a degree of protection in the event of loss of jobs. Labour mobility and ability of enterprises to adjust could also be facilitated by such measures of social protection. Simultaneously, it would be useful to introduce measures to create incentives for new enterprises to come up and existing enterprises to expand. Indeed, it may not be too difficult to find models where a combination of employment security and social protection with incentives for enterprises create an environment of security with flexibility.16

While talking about social protection, one must not forget the informal economy, because that is where vast numbers work in South Asian countries. And many of them are vulnerable to poverty, if not in outright poverty. And recent experience of the ESEA countries during the economic crisis shows that sharp economic downturns can lead to a crowding of workers in the informal sector. In order to avoid seeing them and their family members relapsing into poverty, it is important to have measures of social protection in the informal economy also. While it may not be realistic to have the full range of such measures introduced at once throughout this vast segment of the economy, there are examples to indicate that it is not unrealistic to at least start with some form of social assistance and health benefit.17

Before concluding the discussion on institutional interventions in labour markets, it would be in order to say a few words on the role of the relevant institutions (e.g., the employers’ and workers’ organizations) and of a culture of dialogue between the relevant parties. The experience of the Asian economic crisis has not only shown the negative consequences of a relative neglect of basic labour standards and social protection, but also the potential as well as the actual role of social dialogue in ensuring a smoother adjustment to structural change and in coping with the crisis. In this respect, the contrast between Korea and other countries can be quite illustrative. In Korea, a Tripartite Social Accord was signed soon after the onset of the crisis, and that constituted the strong response to the social costs of crisis, including the extension of the coverage of unemployment insurance and the introduction of various labour market programmes mentioned earlier. In Indonesia and Thailand, where labour movements were weaker, measures to contain the social costs of the crisis were also not as strong (Lee, 1998). Korea, on the other hand, could adjust to and emerge out of the crisis more smoothly than the other countries. Social dialogue can, thus, yield both economic and social benefits.18

16 Sweden provides an example of such model. For details, see International Herald Tribune, 9-10 October 1999. Citing this example should not, of course, be taken to mean that it can be directly emulated by developing countries. Instead, the example is provided here with a view to indicating the existence of alternative models. 17 A joint ILO -World Bank project on micro insurance has demonstrated the feasibility of health insurance for the poor. For a brief report on some such examples, see New York Times, 4 December 2002 (article by Daniel Altiman). See, also, Van Ginneken (1998). 18 Indeed, social dialogue can contribute to adjustment through a variety of means other than simple retrenchment of workers. Such means include agreed reduction in wages and other benefits, furloughing of workers for an agreed period, etc. see Campbell (2001).

30

But meaningful and effective social dialogue in the area of labour policy requires strong labour market institutions that are free of political influence and an enabling political and social environment. It would thus be important for countries to create such environment and strengthen the capacity of labour market institutions so that they can effectively participate in social dialogue and facilitate adjustment to structural changes as well as to sharp economic downturns in an economy.

31

References

Betcherman, Gordon and Rizwanul Islam (2001): East Asian Labour Markets and the Economic Crisis: Impacts, Responses and Lessons. World Bank, Washington, D.C., and International Labour Office, Geneva.

Campell, Duncan (2001): “Social Dialogue and Labour Market Adjustment in East Asia after the Crisis”, in Betcherman and Islam (2001).

Edwards, Alejandra Cox and Chris Manning (2001): “The Economics of Employment Protection and Unemployment Insurance Schemes: Policy Options for Indonesia, Malaysia, Philippines and Thailand”, in Betcherman and Islam (2001).

Freeman, R.B. (1993): “Labour Markets and Institutions in Economic Development”, in American Economic Review, May 1993.

International Labour Office (ILO) (1997): World Employment Report 1996/97. ILO, Geneva.

- (2002a): Labour Market Policies and Poverty Reduction Strategies in Recovery from the Asian Crisis. Report of a Tripartite Sub-Regional Seminar, Jakarta, 29 April – 1 May 2002. ILO, Bangkok.

- (2002b): Employment Policies for Poverty Reduction in Indonesia. Recovery and Reconstruction Department, ILO, Geneva.

Islam, Rizwanul (2001a): “Poverty Alleviation, Employment and Labour Market: Lessons from the Asian Experience and Policies”. Paper presented at the Asia and Pacific Forum on Poverty organized by the Asian Development Bank, 5-9 February 2001, Manila.

- (2001b): “Employment Implications of the Global Economic Slowdown 2001: Responding with a Social Focus”. Issues in Employment and Poverty Discussion Paper 2, Recovery and Reconstruction Department, ILO, Geneva.

- , et al., (2001a): “The Economic Crisis: Labour Market Challenges and Policies in Indonesia”, in Betcherman and Islam (2001).

- , et al., (2001b): “Active Labour Market Policies in East and South-East Asia: What has been done and what can be done”. Paper presented at a joint ILO-World Bank Seminar on Labour Market Policies held in Manila on 1-2 March 2001.

- and J. Krishnamurty (2002): “Employment, Labour Market and Poverty during Economic Recovery and Slowdown in East and South East Asia”. Paper presented at a seminar on Strengthening Employment and Labour Market Policies for Poverty Alleviation and Economic Recovery in East and South East Asia, Jakarta, 29 April – 1 May 2002.

Islam, Iyanatul (2002): “Poverty, Employment and Wages: An Indonesian Perspective”, in ILO (2002a).

Islam, Iyanatul and Anis Chowdhury (2000): The Political Economy of East Asia: Post Crisis Debates. Oxford University Press, Melbourne.

32

Khan, A.R. (1994): “Stabilization, Structural Adjustment and the Labour Market: Issues and Evidence from Asia”, in Rizwanul Islam (ed.). Social Dimensions of Economic Reforms in Asia, ILO, New Delhi. - (2001): “Employment Policies for Poverty Reduction”. Issues in Employment and Poverty Discussion Paper 1. Recovery and Reconstruction Department. ILO, Geneva. Layard, Richard, Stephen Nickell, and Richard Jackman (1994): The Unemployment Crisis. Oxford University Press, New York. Lee, Eddy (1998): The Asian Financial Crisis: The Challenge for Social Policy. ILO, Geneva. Lewis, W.A. (1954): “Economic Development with Unlimited Supplies of Labour”, in Manchester School of Economic and Social Studies, vol. 22, No. 2.

Stiglitz, Joseph E. and Shahid Yusuf (eds.) (2001): Rethinking the East Asian Miracle. World Bank and Oxford University Press, Washington D.C. and Oxford 2001. Sundaram, K. and Suresh Tendulkhar (2002): “The Working Poor in India: Employment-Poverty Linkages and Employment Policy Options”. Issues in Employment and Poverty Discussion Paper 4. Recovery and Reconstruction Department, ILO, Geneva. Van Ginneken, Wouter (ed.) (1998): Social Security for all Indians. Oxford University Press, Delhi. World Bank (1993): The East Asian Miracle: Economic Growth and Public Policy. World Bank, Oxford University Press, 1993.

33

Issues in Employment and Poverty Discussion Papers

1. Azizur Rahman Khan: Employment Policies for Poverty Reduction (November 2001) 2. Rizwan Islam: Employment Implications of the Global Economic Slowdown 2001:

Responding with a Social Focus (November 2001) 3. Rohini Nayyar: The Contribution of Public Works and Other Labour-Based

Infrastructure to Poverty Alleviation: The Indian Experience (September 2002) 4. K. Sundaram and Suresh D. Tendulkar: The Working Poor I India: Employment-

Poverty Linkages and Employment Policy Options 5. Stephen Devereux: From Workfare to Fair Work: The Contribution of Public Works

and Other Labour-based Infrastructure Programmes to Poverty Alleviation 6. Mustafa K. Mujeri: Bangladesh: Bringing Poverty Focus in Rural Infrastructure

Development 7. G.K. Chadha: Rural Employment in India: Current Situation, Challenges and

Potential for Expansion